Enhanced lotus effect optimization algorithm for efficient problem-solving in high-dimensional complex landscapes

Abstract

The lotus effect optimization algorithm (LEOA) is a metaheuristic technique based on the self-cleaning functionality of the lotus flower and other adaptive features. Although LEOA is effective in optimization exercises, it encounters challenges like early stagnation and an imbalanced trade-off between exploration and exploitation. These shortcomings restrict the performance of the algorithm in some complex high-dimensional problem spaces. This paper proposes an enhanced LEOA (ELEOA), in which LEOA is integrated with particle swarm optimization (PSO) to improve convergence rate, stability, and search effective-ness. To validate the performance of ELEOA, it was benchmarked against 12 recent algorithms assessed on a collection of 20 classical benchmarks along with the IEEE CEC 2019 suite. Experimental results demonstrate that ELEOA outperformed competing algorithms on 18 out of 20 classical functions and ranked among the top three in 18 of the 20 CEC 2019 functions. In terms of convergence speed, ELEOA achieved a 35.7 % improvement over the average of all compared algorithms. Furthermore, the algorithm was applied to three engineering design problems, where it achieved optimal or near-optimal solutions with reduced computational cost. These results confirm ELEOA's robustness, accuracy, and potential for solving complex real-world optimization problems.